Peripheral and Central Glucose Utilizations Modulated by Mitochondrial DNA 10398A in Bipolar Disorder

Full text

Psychoneuroendocrinology
(2015)
55,
72—80
Available
online
at
www.sciencedirect.com
ScienceDirect
jou
rn
al
h
om
epage:
www.elsevier.com/locate/psyneuen
Peripheral
and
central
glucose
utilizations
modulated
by
mitochondrial
DNA
10398A
in
bipolar
disorder
Cheng-Ta
Lia,b,
Ya-Mei
Baia,b,
Jen-Chuen
Hsiehc,d,
Hsin-Chen
Leee,
Bang-Hung
Yangf,
Mu-Hong
Chena,b,
Wei-Chen
Lina,b,
Chia-Fen
Tsaia,b,
Pei-Chi
Tu a,b,
Shyh-Jen
Wangf,
Tung-Ping
Sua,b,d,∗
aDepartment
of
Psychiatry,
Taipei
Veterans
General
Hospital,
Taipei,
Taiwan
bDivision
of
Psychiatry,
Faculty
of
Medicine,
National
Yang-Ming
University,
Taipei,
Taiwan
cIntegrated
Brain
Research
Unit,
Division
of
Clinical
Research,
Department
of
Medical
Research,
Taipei
Veterans
General
Hospital,
Taiwan
dInstitute
of
Brain
Science,
National
Yang-Ming
University,
Taipei,
Taiwan
eDepartment
and
Institute
of
Pharmacology,
School
of
Medicine,
National
Yang-Ming
University,
Taipei,
Taiwan
fDepartment
of
Nuclear
Medicine,
Taipei
Veterans
General
Hospital,
Taipei,
Taiwan
Received
6
December
2014;
received
in
revised
form
17
January
2015;
accepted
4
February
2015
KEYWORDS
Bipolar
disorder;
Mitochondria;
Gene;
Polymorphism;
Glucose
homeostasis
Summary
Bipolar
disorder
(BD)
is
highly
heritable
and
associated
with
dysregulation
of
brain
glucose
utilizations
(GU).
The
mitochondrial
DNA
(mtDNA)
10398A
polymorphism,
as
a
reported
BD
risk
factor,
leads
to
deficient
glycolytic
energy
production
by
affecting
mitochondrial
matrix
pH
and
intracellular
calcium
levels.
However,
whether
mtDNA-10398A
has
functional
effects
on
the
brain
and
how
our
body
responds
remain
elusive.
We
compared
peripheral
and
central
glucose-utilizing
patterns
between
mtDNA
A10398G
polymorphisms
in
BD
and
their
unaffected
siblings
(BDsib).
Since
siblings
carry
identical
mtDNA,
we
hypothesized
that
certain
character-
istics
co-segregate
in
BD
families.
We
recruited
twenty-seven
pairs
of
non-diabetic
BD
patients
and
their
BDsib
and
30
well-matched
healthy
control
subjects
(HC).
The
following
were
investi-
gated:
mtDNA,
fasting
plasma
glucose/insulin,
cognitive
functions
including
Montreal
Cognitive
Assessment
(MoCA),
and
brain
GU
at
rest.
Insulin
resistance
was
rechecked
in
sixty-one
sub-
jects
(19-BD,
18-BDsibib,
and
24-HC)
six
months
later.
We
found
that
BD-pairs
(BD
+
BDsib)
carried
more
mtDNA-10398A
and
had
higher
fasting
glucose,
even
after
controlling
for
many
∗Corresponding
author
at:
No.
201,
Sec.
2,
Shih-Pai
Road,
Beitou,
Taipei
112,
Taiwan.
Tel.:
+886
2
28757778.
E-mail
addresses:
tpsu@vghtpe.gov.tw,
tomsu0402@gmail.com
(T.-P.
Su).
http://dx.doi.org/10.1016/j.psyneuen.2015.02.003
0306-4530/©
2015
Elsevier
Ltd.
All
rights
reserved.

Mitochondrial
DNA
and
glucose
homeostasis
in
BD
73
covariates.
BD-pairs
had
abnormally
lower
dorso-prefrontal-GU
and
higher
cerebellar-GU,
but
only
BD
demonstrated
lower
medio-prefrontal-GU
and
MoCA.
Subjects
carrying
mtDNA-10398A
had
significantly
lower
prefrontal-GU
(FWE-corrected
p
<
0.05).
An
abnormal
inverse
pattern
of
insulin-GU
and
insulin-MoCA
correlation
was
found
in
BD-pairs.
The
insulin-MoCA
correlation
was
particularly
prominent
in
those
carrying
mtDNA-10398A.
mtDNA-10398A
predicted
insulin
resistance
6
months
later.
In
conclusion,
mtDNA-10398A
was
associated
with
impaired
prefrontal-
GU.
An
up-regulation
of
glucose
utilizations
was
found
in
BD-pairs,
probably
compensating
for
mtDNA-10398A-related
energy
loss.
©
2015
Elsevier
Ltd.
All
rights
reserved.
1.
Introduction
Bipolar
disorder
(BD)
is
a
highly
heritable
disorder
(McGuffin
et
al.,
2003),
characterized
by
mood
swings
between
extremely
high-energy
(mania)
and
low-energy
(depression)
states.
A
wide
range
of
drugs
available
for
the
treatment
of
BD
(e.g.,
valproate,
lithium,
and
atypical
antipsychotics)
are
able
to
treat
mood
symptoms
(Bai
et
al.,
2013),
leaving
some
other
symptoms
(e.g.,
cognitive
dysfunction)
untreated.
Such
insufficient
therapeutic
effects
of
the
available
med-
ications
imply
that
the
underlying
pathophysiology
of
BD
remains
obscure
and
new
therapeutic
target
is
warranted.
Several
lines
of
evidence
indicate
that
mitochondrial
dysfunction
plays
a
critical
role
in
the
pathophysiology
of
BD
(Kato,
2006;
de
Sousa
et
al.,
2014).
For
example,
studies
investigating
in
vivo
brain
chemistry
by
magnetic
resonance
spectroscopy
(MRS)
revealed
that
BD
involves
impaired
oxidative
phosphorylation,
altered
phospholipid
metabolism,
increased
intracellular
pH,
a
resultant
shift
toward
glycolytic
energy
production,
and
a
decrease
in
total
cerebral
energy
production
(Stork
and
Renshaw,
2005).
Mitochondria
are
the
most
important
organelles
in
most
eukaryotic
cells
that
regulate
cellular
metabolism
and
con-
vert
glucose
to
ATP
energy
which
organisms
need.
Glucose
is
the
main
energy
source
for
the
human
brain.
Impaired
brain
glucose
utilizations
have
been
reported
in
functional
studies
of
positron
emission
tomography
(PET)
investigat-
ing
BD
patients,
such
as
hypo-utilizations
in
the
prefrontal
cortex
(Buchsbaum
et
al.,
1986;
Baxter
et
al.,
1989).
The
observed
hypofrontality
was
left
untreated
even
dur-
ing
patients’
euthymic
state
and
also
accountable
for
the
cognitive
deficits
in
remission
(Li
et
al.,
2012).
Impaired
cognitive
function
(e.g.,
executive
dysfunction)
has
been
also
reported
in
the
unaffected
relatives
of
BD
patients
(Arts
et
al.,
2008;
Balanza-Martinez
et
al.,
2008;
Bora
et
al.,
2009).
However,
whether
the
mitochondrial
dysfunc-
tion
hypothesis
of
BD
could
also
explain
the
altered
cerebral
glucose
utilizations
(e.g.,
impaired
prefrontal
utilization)
remains
elusive.
Another
piece
of
evidence
supporting
BD
as
a
heritable
disorder
is
from
mitochondrial
genetic
studies.
For
exam-
ple,
mitochondria
DNA
(mtDNA)
10398A
polymorphism
is
one
of
the
reported
mtDNA
polymorphisms
associated
with
BD
(Kato
and
Kato,
2000).
BD
patients
carrying
mtDNA-
10398A
seems
to
respond
better
to
lithium
(Washizuka
et
al.,
2003).
Given
the
polymorphism
plays
a
role
in
glycolytic
energy
production
by
affecting
mitochondrial
matrix
pH
and
intracellular
calcium
levels,
the
resulting
mitochondrial
dys-
function
would
lead
to
an
ineffective
production
of
ATP
energy
from
glucose
and
might
have
an
impact
on
human
brains.
However,
whether
mtDNA-10398A
has
functional
effects
on
brain
glucose
utilizations
(GU)
and
long-term
glu-
cose
homeostasis
remains
elusive.
Therefore,
in
the
present
study,
we
investigated
mtDNA
A10398G
polymorphism,
fasting
plasma
glucose
and
insulin,
cognitive
functions,
and
brain
GU
as
measured
by
res-
ting 18F-FDG
PET,
and
their
relationships
among
three
groups
of
subjects
without
a
history
of
diabetes
melli-
tus:
(1)
BD
patients,
(2)
unaffected
siblings
of
the
BD
patients
(BDsib),
and
(3)
healthy
control
subjects
(HC).
We
hypothesized
that,
in
response
to
insufficient
glycolytic
energy
production,
some
compensatory
glucose-utilizing
characteristics
might
be
observed
and
co-segregate
in
BD
families.
Since
siblings
carry
identical
mtDNA,
we
thus
investi-
gated
central
and
peripheral
glucose-utilizing
patterns
in
BD
and
BDsib
and
the
effects
from
mtDNA-10398
polymorphisms
were
compared.
To
further
investigate
whether
long-term
mitochondrial
dysfunction
might
affect
the
glucose
homeo-
stasis,
homeostasis
model
assessment
for
insulin
resistance
(HOMA-IR)
(Matthews
et
al.,
1985)
was
tested
at
baseline
and
re-tested
six
months
later
between
subjects
with
and
without
the
mtDNA-10398A.
2.
Methods
and
materials
2.1.
Participants
Eighty-four
subjects
were
recruited,
including
twenty-seven
pairs
of
stable
BD
subtype
I
patients
and
their
unaffected
BDsib
(age
differences
<
5
years
in
the
BD
pairs)
and
30
age-,
gender-,
and
ethnicity-matched
HC
(Table
1).
Subtype
I
of
BD
patients
were
selected
mainly
due
to
its
high
heritability
(Gershon
et
al.,
1982),
and
the
diagnoses
were
established
by
structured
history-taking
and
administration
of
the
Mini
International
Neuropsychiatric
Interview
(MINI)
based
on
the
Fourth
Edition
of
the
Diagnostic
and
Statistical
Manual
sys-
tem
criteria.
The
recruited
BDsib
and
HC
were
all
free
of
axis
I
disorders,
and
all
subjects
were
free
of
major
medi-
cal
and
neurological
illness,
a
lifetime
alcohol
or
substance
abuse
history.
HC
subjects
with
a
family
history
of
an
axis
I
disorder
were
also
excluded.
To
reduce
potential
confound-
ing
effects
from
medications,
all
recruited
patients
were
either
clinically
stable
off
medications
or
had
stop
taking
medications
for
at
least
five
days
before
the
enrollment.
Fur-
thermore,
sixty-one
subjects
(19-BD,
18-BDsib,
and
24-HC)
had
fasting
plasma
sugar
and
insulin
rechecked
six
months

74
C.-T.
Li
et
al.
Table
1
Demographic
data,
clinical
variables,
and
mitochondrial
DNA
(mtDNA)
polymorphisms
related
to
bipolar
disorders
(BD)
among
three
age-,
gender-,
education-,
and
ethnicity-matched
groups:
BD,
unaffected
siblings
of
BD
patients
(BDsib),
and
healthy
control
subjects
(HC).
BD
(n
=
27)
BDsib
(n
=
27)
HC
(n
=
30)
F/t
value
p
Demographic
data
Gender
(M/F) 15/12
13/14
16/14
2=
0.314
0.855
Age
(y/o) 39.7
(9.8) 39.6
(11.1) 39.0
(10.6) 0.044
0.957
Education
(N/N/N)a13/11/3
7/17/3
9/19/2
2=
4.25 0.373
BMI
25.7
(4.5)
24.1
(2.6)
22.6
(3.9)
1.218
0.304
Clinical
variables
HDRS-17
6.8
(6.4)
1.1
(1.8)
0.8
(1.4)
15.609
0.001*
YMRS
3.5
(3.5)
0.6
(1.7)
0.1
(0.2)
10.335
0.001*
Duration
of
illness
(years)
13.4
(9.3)
0
0
—
—
Manic
episodes
(number)
5.1
(5.3)
0
0
—
—
Major
depressive
episodes
(number)
4.1
(3.9)
0
0
—
—
Hospitalizations
(number)
2.2
(2.7)
0
0
—
—
Mitochondria
DNA
(mtDNA)
gene
polymorphism
mtDNA
10398
A/G
20/7
20/7
14/16
6.310
0.043*
Note:
HDRS-17,
17-item
Hamilton
Depression
Rating
Scale;
YMRS,
Young
Mania
Rating
Scale;
BMI,
Body
Mass
Index
[=weight
(kg)/height2
(m2)].
*p
<
0.05.
aN/N/N:
≤12
years/12—16
years/
>
16
years
of
education.
after
the
recruitment.
Other
details
please
refer
to
the
sup-
plementary
material.
The
study
was
performed
in
accordance
with
the
Decla-
ration
of
Helsinki
and
was
approved
by
the
Ethics
Review
Committee
of
Taipei
Veterans
General
Hospital
(2013-06-
028B).
2.2.
DNA
extraction
and
analysis
for
mtDNA
polymorphisms
Blood
samples
were
collected
by
venipuncture
into
EDTA-
2Na
vacutubes.
Total
DNA
from
peripheral
blood
cells
was
extracted
with
the
QIAamp
DNA
Mini
kit
(QIAGEN)
accord-
ing
to
the
manufacturer’s
instructions.
The
final
DNA
was
dissolved
in
doubly
distilled
water
and
frozen
at
−30 ◦C
until
use.
The
mtDNA
polymorphism
site
at
A10398G
was
genotyped
using
the
polymerase
chain
reaction—restriction
fragment
length
polymorphism
(PCR-RFLP)
method.
Other
details
please
refer
to
the
supplementary
material.
2.3.
Fasting
plasma
glucose
(FPG),
insulin,
and
HOMA-IR
The
fasting
blood
samples
for
glucose
and
insulin
were
col-
lected
from
8:00
to
10:00
am.
Subjects
fasted
for
at
least
8
h
before
the
blood
exam.
Insulin
was
assayed
by
enzyme-
linked
immunosorbent
assay
(ELISA)
kits
(Millipore,
Billerica,
MA,
USA).
The
lowest
level
of
insulin
that
can
be
detected
by
this
assay
is
1
␮U/ml.
The
assay
was
performed
fol-
lowing
the
vendor’s
instructions.
The
final
absorbance
of
the
mixture
was
measured
and
analyzed
at
450
nm
using
an
ELISA
plate
reader
by
Bio-Tek
PowerWave
XS
and
Bio-
Tek’s
KC
junior
software
(Winooski,
VT,
USA).
A
linear
regression
R-square
value
of
0.95
and
above
was
a
reliable
standard
curve.
Homeostasis
model
assessment
for
insulin
resistance
(HOMA-IR)
was
calculated
using
the
following
for-
mula:
HOMA-IR
=
FPG
(mg/dl)
×
fasting
insulin
(␮U/ml)/405
(Matthews
et
al.,
1985).
Higher
HOMA-IR
values
represented
greater
degrees
of
insulin
resistance.
2.4.
Cognitive
measurement
The
Wisconsin
card
sorting
test
(WCST)
(Heaton
et
al.,
1993)
was
adopted
for
evaluating
executive
functions
which
bet-
ter
reflected
activity
of
the
prefrontal
cortex
(PFC)
(Li
et
al.,
2012).
In
addition,
the
Mini-Mental
State
Examina-
tion
(MMSE)
(Folstein
et
al.,
1975)
and
Montreal
Cognitive
Assessment
(MoCA)
(Nasreddine
et
al.,
2005)
were
used
to
detect
global
cognitive
functions
in
BD
patients.
While
the
MoCA
was
more
sensitive
than
the
MMSE
in
detecting
mild
cognitive
impairment
(Nasreddine
et
al.,
2005),
both
demonstrated
good
agreement
with
the
executive
function
in
the
subjects
with
cognitive
impairments
(e.g.,
post-
stroke)
(Cumming
et
al.,
2013).
Additional
details
were
available
in
the
online
supplemental
materials.
2.5.
Imaging
studies
2.5.1.
Structural
magnetic
resonance
imaging
(MRI)
MR
images
were
acquired
with
a
3.0
GE
Discovery
750
whole-
body
high-speed
imaging
device.
High-resolution
structural
T1-weighted
images
were
acquired,
for
improving
co-
registration
of
the
PET
images,
in
the
sagittal
plane
using
a
high
resolution
sequence
(repetition
time
(TR),
2530
ms;
echo
spacing,
7.25
ms;
echo
time
(TE),
3
ms;
flip
angle
7◦)
with
isotropic
1
mm
voxels
and
FOV
=
256
mm
×
256
mm.

Mitochondrial
DNA
and
glucose
homeostasis
in
BD
75
2.5.2.
Positron
emission
tomography
(PET)
2.5.2.1.
PET
data
acquisition. 18F-FDG
PET
scans
of
at
rest
glucose
utilizations
were
acquired
on
a
PET/CT
scan-
ner
(Discovery
VCT;
GE
Healthcare,
USA)
with
the
3D
brain
mode.
All
subjects,
except
three
subjects
in
the
HC,
com-
pleted
the
PET
scans.
In
total,
81
subjects
(27-BD,
27-BDsib,
and
27-HC)
were
included
in
the
PET
analysis.
All
subjects
fasted
for
at
least
8
h
before
the
PET
examination.
PET
images
were
acquired
within
45
min
after
an
intravenous
injection
of
approximately
370
MBq
per
70
kg
of 18F-FDG
(adjusted
by
their
weight).
Other
details
of
acquisitions
were
identical
and
mentioned
in
our
previous
study
(Li
et
al.,
2012).
2.5.2.2.
PET
analysis.
PET
data
were
analyzed
using
Sta-
tistical
Parametric
Mapping
version
8
software
(SPM8;
Wellcome
Department
of
Cognitive
Neurology,
Institute
of
Neurology,
University
College
London,
London,
England)
implemented
in
Matlab
7.1
(The
Mathworks
Inc.,
Sherborn,
MA,
USA).
A
group-specific
MRI-aided 18F-FDG
template
was
created
(Signorini
et
al.,
1999;
Gispert
et
al.,
2003)
and
used
to
normalize
each
subject’s
PET
images,
followed
by
smoothing
with
a
3D
Gaussian
kernel
(FWHM
=
8
mm).
Other
details
please
referred
to
the
online
supplemental
materials.
We
first
calculated
and
compared
global
utilizations
among
the
three
groups.
Then,
the
overall
grand
mean
from
the
PET
scans
was
centered
and
normalized
to
100,
and
global
variance
across
the
scans
was
removed
by
analysis
of
covariance
(ANCOVA)
(Lee
et
al.,
2008).
To
assess
potential
group
differences
of
normalized
brain
GU,
an
ANCOVA
with
age,
gender
(Willis
et
al.,
2002),
mood
symptoms,
BMI,
and
global
gray
matter
values
as
covariates
of
no
interest
was
used
for
between-group
com-
parisons.
The
significance
thresholds
for
3-group
comparison
(F-map)
were
set
at
a
voxel-level
p
<
0.05
[FWE(family-wise
errors)
−
corrected
for
multiple
comparisons].
We
further
inspected
the
between-group
differences
(BD
vs.
HC;
BDsib
vs.
HC;
BD
vs.
BDsib)
on
the
identified
F-map,
and
the
significance
threshold
for
the
subsequent
between-group
t-test
analyses
was
set
at
p
<
0.0167
(=0.05/3,
corrected
for
3
group
comparisons).
A
two-way
ANCOVA
with
mtDNA
and
group
as
the
two
independent
factors
was
used
to
compare
GU
between
subjects
carrying
mtDNA-10398A
and
mtDNA-10398G
and
interactions
between
mtDNA
and
group.
Since
a
priori
knowledge
about
the
abnormal
brain
net-
work
of
BD
involved
brain
regions
such
as
PFC
(Buchsbaum
et
al.,
1986;
Baxter
et
al.,
1989;
Li
et
al.,
2012),
we
aimed
to
explore
potential
influences
of
mtDNA
on
brain
GU
by
setting
the
significance
thresholds
at
voxel-level
uncor-
rected
p
<
0.001
and
cluster-level
corrected
p
<
0.05.
Here
we
focused
on
medial
PFC
and
dorsolateral
PFC,
since
these
PFC
subregions
were
the
most
significant
areas
found
to
be
hypoactive
in
euthymic
BD
patients
(Li
et
al.,
2012).
Finally,
multivariate
regression
analyses,
with
controlling
for
age,
sex,
mood
symptoms,
BMI,
and
global
glucose
utiliza-
tions,
were
performed
to
study
the
correlations
between
insulin
(i.e.,
the
predictor)
and
brain
GU
(i.e.,
the
outcome
variable)
in
each
group.
Then,
a
t-test
was
used
to
test
whether
the
slope
of
a
regression
line
differs
significantly
from
0.
A
cluster-level
corrected
p-value
less
than
0.05
in
the
correlation
analyses
was
deemed
to
be
statistically
significant.
2.6.
Statistical
analysis
for
other
variables
For
analysis
of
the
demographic
data
and
clinical
variables,
SPSS
16.0
software
(SPSS
Inc.,
Chicago,
IL)
was
used.
One-
way
ANOVA
(or
Student’s
t
tests)
and
chi-square
tests
were
applied
to
compare
the
continuous
and
categorical
variables
among
groups,
respectively.
Pearson’s
correlation
was
used
to
investigate
relationships
between
continuous
variables
(e.g.,
insulin
and
cognitive
function).
A
p-value
less
than
0.05
was
considered
statistically
significant.
3.
Results
3.1.
Demographic
data,
clinical
variables
and
mtDNA
gene
polymorphism
There
was
no
difference
in
the
age,
gender,
educational
lev-
els,
and
BMI
among
the
three
groups.
BD
subjects
were
in
clinically
stable
condition
but
had
statistically
more
severe
mood
symptoms
(i.e.,
HDRS-17
and
YMRS
scores)
(Table
1).
Most
of
the
patients
were
in
remission
at
baseline
(i.e.,
81.8%
of
all
patients
had
a
score
of
≤7
on
the
HDRS-17
and
≤7
on
the
YMRS
for
at
least
two
weeks),
while
89.4%
of
the
patients
at
the
6th-month
follow-up
(n
=
17)
were
also
euthymic.
The
recruited
BD
patients
ranged
widely
in
the
distribution
of
the
duration
of
illness
(1—35
years),
with
a
mean
(SD)
of
13.4
(9.3)
years
(Table
1).
The
distribution
of
the
mtDNA-10398A,
the
known
mitochondrial
risk
polymor-
phism
gene,
was
statistically
significant
among
three
groups
(p
=
0.043)
(Table
1).
More
specifically,
the
mtDNA-10398A
was
as
frequent
in
the
BD
patients
and
the
BDsib
subjects
and
was
more
frequent
in
the
BD
pairs
than
in
the
HC.
3.2.
Fasting
plasma
glucose
(FPG),
insulin,
HOMA-IR,
and
cognitive
functions
The
BD
pairs,
including
BD
and
BDsib,
had
significantly
higher
FPG
than
the
HC,
even
after
controlling
out
covariates
such
as
age
and
BMI
(Table
2)
(Fig.
S1A).
Direct
comparisons
among
three
groups
demonstrated
that
the
BD
had
a
higher
HOMA-IR
level
than
the
HC
(Table
2)
(Fig.
S2).
However,
after
adjusting
for
covariates,
no
significant
differences
existed
in
the
insulin
and
HOMA-IR.
The
BD
cognitive
performance
was
significantly
worse
than
the
BDsib
and
HC
on
many
domains
including
the
MMSE,
MoCA,
and
the
executive
functions
as
measured
by
WCST.
However,
after
adjusting
for
covariates,
BD
performed
significantly
worse
than
the
other
two
groups
only
on
the
MoCA
(Table
2)
(Fig.
S1B).
Since
BMI
is
a
well-known
factor
for
glucose
levels,
we
excluded
subjects
who
were
severely
underweight
(BMI
<
18.5)
and
obese
(BMI
>
26)
in
the
subsequent
analy-
ses.
If
the
analyses
were
done
in
all
the
subjects
(n
=
50,
BD/BDsib/HC
=
15/18/17),
subjects
with
mtDNA-10398A
had
non-significantly
higher
levels
of
insulin,
HOMA-IR,
and
FPG
than
those
carrying
mtDNA-10398G
after
controlling
for
age,
sex,
education,
mood
symptoms,
and
BMI.
However,
if
the
analysis
was
done
in
separate
groups,
only
the
BD
group
demonstrated
that
subjects
with
mtDNA-10398A
had
higher
levels
of
insulin
(p
=
0.040),
HOMA-IR
(p
=
0.029),
and

76
C.-T.
Li
et
al.
Table
2
Demographic
data,
clinical
variables,
and
mitochondrial
DNA
(mtDNA)
polymorphisms
related
to
bipolar
disorders
(BD)
among
three
age-,
gender-,
education-,
and
ethnicity-matched
groups:
BD,
unaffected
siblings
of
BD
patients
(BDsib),
and
healthy
control
subjects
(HC).
BD
BDsib
HCs
ANOVA
F-value
ANOVA
p-value
ANCOVAap-value
Post
hoc
(LSD)
Glucose
homeostasis
Glucose
(mg/dl)
89.9
(9.5)
90.3
(11.9)
82.0
(11.0)
5.401
0.006*0.023*BD
>
HC,
BDsib
>
HC
Insulin
(␮U/ml)
6.4
(6.2)
4.9
(2.8)
3.9
(3.4)
2.386
0.092
0.815
—
HOMA-IR
1.4
(1.3)
1.1
(0.7)
0.8
(0.7)
3.573
0.033*0.563
—
General
cognition
MMSE
27.9
(1.4)
29.3
(1.2)
29.4
(0.8)
14.485
0.000*0.190
—
MoCA
25.3
(2.5)
28.2
(1.5)
28.1
(1.5)
21.758
0.000*0.019*BD
<
BDsib,
BD
<
HC
Executive
functions
—
Wisconsin
Card
Sorting
Test
%errors
28.0
(13.8)
19.4
(8.7)
20.1
(13.5)
4.182
0.019*0.229
—
%Conceptual
level
responses
64.4
(19.3)
77.1
(12.2)
75.5
(18.5)
4.492
0.014*0.192
—
Categories
completed
4.8
(2.0)
5.7
(1.1)
5.6
(1.4)
2.541
0.085
0.318
—
Note:
HOMA-IR,
Homeostasis
Model
Assessment
of
Insulin
Resistance;
MMSE,
Mini-Mental
State
Examination;
MoCA,
the
Montreal
Cognitive
Assessment.
*p
<
0.05.
aANCOVA
analysis
with
age,
sex,
education,
mood
symptoms,
and
BMI
as
covariates
of
no
interests.
FPG
(p
=
0.076,
in
a
trend
significance)
than
those
carrying
mtDNA-10398G
(Table
S1).
No
statistical
significance
existed
on
cognitive
functions.
3.3.
Regional
GU
and
correlations
with
insulin
If
analyzed
in
all
study
subjects,
subjects
with
mtDNA-
10398A
polymorphism
had
lower
GU
than
those
carrying
mtDNA-10398G
in
a
big
cluster
(3960
voxels,
peak
=
[16,
28,
50],
t
=
4.30)
covering
the
mPFC
and
dPFC
(boxed
in
red,
Fig.
1).
Subsequent
analyses
showed
that
the
finding
of
lower
mPFC-
and
dPFC-GU
associated
with
mtDNA-
10398A
existed
for
subjects
across
BD,
BDsib,
or
HC
groups
(Fig.
1).
No
interaction
between
group
and
mtDNA
was
found.
The
global
GU
among
the
three
groups
was
not
statistically
significant
(F
=
1.264,
p
=
0.295).
The
ANCOVA
analysis
demonstrated
that
the
most
significant
brain
regions
among
3
groups
included
two
large
clusters
[PFC
(1091
voxels,
peak
MNI
coordinates
[x,
y,
z]
=
[−26,
54,
18],
F
=
8.27)
and
cerebellum
(3659
voxels,
peak
=
[−20,
48,
28],
F
=
9.11)]
(voxel-level
p
<
0.05,
FWE-corrected
for
multi-
ple
comparisons)
(boxed
in
red,
Fig.
S3).
These
clusters
covered
medial
PFC
(mPFC),
dorsal
PFC
(dPFC),
ante-
rior
cingulate
cortex
(ACC),
cerebellar
vermis
and
anterior
lobe
and
posterior
lobe
of
cerebellum.
The
subsequent
between-group
analysis
revealed
that
BD
pairs
all
had
lower
GU
in
the
dPFC
(BD
<
BDsib
<
HC)
and
higher
GU
in
the
cerebellum
(BD
>
BDsib
>
HC)
(Fig.
S3).
A
contrasting
pattern
between
BD
and
BDsib
were
in
the
mPFC,
in
which
BD
had
lower
GU
and
BDsib
had
higher
GU
than
HC
(Fig.
S3).
Regarding
the
insulin-brain
GU
correlations,
the
BD
pairs
all
demonstrated
a
significantly
negative
correlation
in
the
mPFC
and
dPFC
(cluster-level
corrected
p
<
0.05)
(marked
in
solid
arrows,
Fig.
2).
No
such
correlations
existed
in
the
HC.
On
the
other
hand,
BDsib
and
HC
demonstrated
a
similar
pattern
of
a
positive
correlation
between
insulin
and
cere-
bellar
GU
(Fig.
2).
However,
BD
failed
to
demonstrate
this
correlation
(Fig.
2,
in
a
dashed
line),
but
instead
showed
a
negative
correlation
in
the
cerebellum.
3.4.
Insulin-MoCA
correlations
in
those
carrying
mtDNA-10398A
and
mtDNA-10398G
We
compared
the
correlations
among
the
three
groups
with
well-matched
BMI,
by
excluding
subjects
who
were
severely
underweight
(BMI
<
18.5)
and
obese
(BMI
>
26).
The
results
demonstrated
that
HC
had
significantly
negative
correlations
of
insulin
and
MoCA,
whether
they
carried
mtDNA-10398A
(r
=
−0.776,
p
<
0.05)
or
10398G
(r
=
−0.727,
p
<
0.05)
(Fig.
3).
In
contrast,
such
negative
correlations
disappeared
in
BDsib
and
BD,
especially
in
those
carry-
ing
mtDNA-10398A.
BD
patients
carrying
mtDNA-10398A
had
a
strong
positive
correlation
between
insulin
and
MoCA
(r
=
0.526,
p
<
0.05)
(Fig.
3).
Insulin
and
related
glucose
metabolism
may
have
a
different
role
in
BD
families
as
com-
pared
to
HC.
Higher
insulin
in
HC
was
related
to
worse
global
cognition,
while
higher
insulin
in
BD
pairs
was
more
associ-
ated
with
better
cognitive
functions
(Fig.
3).

Mitochondrial
DNA
and
glucose
homeostasis
in
BD
77
Figure
1
Subjects
with
mitochondrial
DNA
10398A
(mtDNA-10398A)
—
had
lower
glucose
utilizations
(GU)
in
the
medical
prefrontal
cortex
(mPFC)
and
dorsal
prefrontal
cortex
(dPFC)
than
those
with
mtDNA-10398G
(cluster-level
corrected
p
<
0.05)
whether
analyzed
in
all
subjects
(boxed
in
red)
or
in
each
group.
A
two-way
ANCOVA
with
mtDNA
and
group
as
the
two
independent
factors
was
used
to
compare
GU
between
subjects
carrying
mtDNA-10398A
and
mtDNA-10398G.
(For
interpretation
of
the
references
to
color
in
figure
legend,
the
reader
is
referred
to
the
web
version
of
the
article.)
Figure
2
Correlations
between
insulin
and
brain
glucose
utilizations
(GU)
in
the
three
groups
(HC,
BDsib,
and
BD).
BD
pairs
all
demonstrated
a
significantly
negative
correlation
in
the
mPFC
and
dPFC
(voxel-level
uncorrected
p
<
0.001
and
followed
by
cluster-
level
corrected
p
<
0.05).
On
the
other
hand,
BDsib
demonstrated
a
similar
pattern
as
HC
with
a
positive
correlation
between
insulin
and
cerebellar
GU.
A
multivariate
regression
analysis
was
performed
to
study
the
correlations
between
insulin
(i.e.,
the
predictor
in
the
regression)
and
brain
GU
(i.e.,
the
outcome
variable
in
the
regression)
in
each
group.
Then,
the
Student’s
t-test
was
used
to
test
whether
the
slope
of
a
regression
line
differs
significantly
from
0.
(For
interpretation
of
the
references
to
color
in
figure
legend,
the
reader
is
referred
to
the
web
version
of
the
article.)
3.5.
mtDNA
A10398G
polymorphism
and
long-term
insulin
resistance
Linear
regression
adopting
age,
sex,
BMI,
FPG
at
base-
line,
fasting
insulin
at
baseline,
and
mtDNA
as
independent
variables
revealed
that
mtDNA
A10398G
polymorphism
(beta
=
−0.240,
t
=
−2.125,
p
=
0.038)
and
BMI
(beta
=
0.400,
t
=
3.159,
p
=
0.003)
predicted
the
HOMA-IR
at
month-6th.
Age
(beta
=
−0.221,
t
=
−1.769,
p
=
0.083),
fasting
insulin
(beta
=
0.236,
t
=
1.862,
p
=
0.068),
fasting
plasma
glucose
(beta
=
−0.087,
t
=
−0.649,
p
=
0.519),
and
sex
(beta
=
0.083,
t
=
0.755,
p
=
0.454)
had
no
predicting
values.
If
the
group
of
the
BD-pairs
was
entered
as
another
variable
in
the
regression
analysis,
mtDNA
A10398G
still
predicted
the
6th-month
HOMA-IR
(beta
=
−0.234,
t
=
−2.016,
p
=
0.048),
but
the
group
(beta
=
0.030,
t
=
0.240,
p
=
0.812)
failed
to
predict.
4.
Discussion
The
strengths
of
the
study
included
a
combination
of
mtDNA,
fasting
insulin/glucose,
brain
glucose
utilizations
(GU),
and
cognitive
functions
to
directly
investigate
the
role
of
mito-
chondrial
dysfunction
in
the
pathophysiology
of
BD,
allowing
comparison
of
the
findings
among
BD,
BDsib,
and
HC
to
study
potential
endophenotypes.
It
is
notable
that
mtDNA
is
maternally
inherited,
so
siblings
have
identical
mDNA.
The
most
important
finding
was
that
mtDNA-10398A
had
functional
effects
on
brain
GU.
The
mtDNA-10398A
polymor-
phism
was
reported
to
affect
mitochondrial
calcium
levels

78
C.-T.
Li
et
al.
Figure
3
Correlations
of
fasting
insulin
and
MoCA
in
subjects
carrying
mtDNA-10398A
(circles)
and
mtDNA-10398G
(squares).
HC
had
significantly
negative
correlations
of
insulin
and
MoCA,
for
both
polymorphisms
mtDNA-10398A
(r
=
−0.776)
and
10398G
(r
=
−0.727).
In
contrast,
such
negative
correlations
disappeared
in
BDsib
and
BD,
especially
in
those
carrying
mtDNA-10398A.
BD
patients
carrying
mtDNA-10398A
had
a
strong
positive
correlation
between
insulin
and
MoCA
(r
=
0.526).
Pearson’s
correlation
was
used
to
investigate
relationships
between
insulin
and
MoCA.
A
p-value
less
than
0.05
was
considered
statistically
significant.
An
asterisk
(*)
represents
p
<
0.05,
and
asterisks
(**)
represents
p
<
0.01.
(For
interpretation
of
the
references
to
color
in
figure
legend,
the
reader
is
referred
to
the
web
version
of
the
article.)
by
a
cybrid
assay
and
altered
calcium
signaling
has
been
reported
in
the
peripheral
blood
cells
of
BD
patients
(Kato
and
Kato,
2000;
Kato
et
al.,
2003).
These
findings
suggested
that
the
polymorphism
could
affect
energy
metabolism.
For
the
first
time,
we
demonstrated
that
subjects
with
mtDNA-
10398A
had
lower
brain
GU
in
the
mPFC
and
dPFC.
The
second
important
finding
was
that
the
BD
and
BDsib
shared
several
trait-like
glucose-utilizing
abnormalities.
The
common
abnormal
features
in
the
BD
pairs
included
(1)
a
significantly
higher
prevalence
of
mtDNA-10398A
which
was
found
to
be
associated
with
lower
GU
in
PFC,
(2)
lower
GU
in
dPFC,
(3)
an
up-regulation
of
fasting
plasma
glucose,
(4)
the
correlations
of
lower
PFC
GU
and
higher
fasting
insulin,
and
(5)
a
‘de
novo’
pattern
of
positive
correlations
between
fasting
insulin
and
PFC
GU
and
between
insulin
and
MoCA
(esp.
in
BD
carrying
mtDNA-10398A).
These
findings
repli-
cated
the
findings
of
the
decreased
PFC
GU
in
a
separate
cohort
of
BD
(Li
et
al.,
2012),
and
were
also
in
line
with
a
large
body
of
evidence
proposing
a
critical
role
of
mito-
chondrial
dysfunction
involved
in
the
pathophysiology
of
BD
(Stork
and
Renshaw,
2005;
Kato,
2006).
Since
mitochondrial
dysfunction
would
lead
to
an
ineffec-
tive
production
of
ATP
energy
from
glucose,
the
observed
increases
of
the
basal
plasma
glucose
levels
in
the
BD
pairs
might
act
as
a
compensatory
mechanism
for
the
mitochondria-related
energy
loss,
including
insufficient
CNS
energy
supply
to
the
brain.
The
supportive
evidence
was
the
presence
of
the
positive
insulin-PFC
GU
and
insulin-MoCA
in
the
BD
pairs.
In
the
BD
pairs,
higher
fasting
insulin
was
associated
with
better
MoCA
(particularly
in
those
with
more
mitochondrial
deficits
—
i.e.,
those
with
mtDNA-10398A),
in
contrast
to
the
negative
insulin-MoCA
finding
in
the
HC.
In
healthy
subjects
without
diabetes
and
dementia,
higher
insulin
levels
were
not
beneficial
and
had
been
consis-
tently
reported
to
be
associated
with
more
cognitive
decline
(Young
et
al.,
2006).
However,
in
subjects
with
impaired
glucose
tolerance
or
insulin
resistance,
insulin
stimulates
‘global’
cerebral
GU
(Hirvonen
et
al.,
2011).
We
speculated
that,
in
order
to
compensate
for
the
impaired
GU
related
to
mtDNA-10398A,
brains
of
subjects
with
BD
diathesis
may
become
insulin-dependent.
Energy
utilized
by
brain
cells
such
as
neurons
and
glia
cells
is
generated
from
plasma
glucose
in
the
mitochondria
and
the
process
needs
insulin.
Insulin
plays
an
important
role
in
transporting
extra-cellular
glucose
into
cells,
and
basal
insulin
facilitates
global
GU
including
cortical
regions
and
cerebellum
(Bingham
et
al.,
2002).
However,
whether
other
mtDNA
genes
play
a
role
in
the
regulation
of
brain
GU
remains
to
be
determined.
These
findings
on
the
compensatory
up-regulation
in
the
peripheral
glucose
level
and
central
GU
co-segregate
within
BD
and
BDsib
and
may
thus
serve
as
an
endophenotype
for
BD.
Furthermore,
we
speculated
that
the
maintained
cognitive
function
could
be
a
result
from
compensatory
up-
regulation
of
brain
GU
by
increasing
glucose
levels
that
we
found
in
the
BD
pairs
(Table
2).
It
has
been
reported
that
when
brains
desperately
need
energy,
e.g.,
during
extreme
fasting
states,
the
strongest
effect
of
insulin
on
brain
glucose
utilizations
is
in
the
cerebellum
(Bingham
et
al.,
2002).
The

Mitochondrial
DNA
and
glucose
homeostasis
in
BD
79
common
finding
of
the
abnormally
increased
cerebellar
GU
in
the
BD
pairs
(Fig.
S3,
BDsib
>
HC
and
BD
>
HC)
could
thus
be
regarded
as
a
reflection
of
the
compensatory
result
from
long-term
insufficient
CNS
glucose
supply
to
the
brains
of
BD
pairs.
Our
result
was
not
only
in
line
with
previous
stud-
ies
showing
cerebellar
abnormalities
in
BD
patients
(Womer
et
al.,
2009),
but
also
in
line
with
structural
and
neuro-
chemical
studies
revealing
that
unaffected
relatives
of
BD
had
enlarged
cerebellum
(Kempton
et
al.,
2009)
and
rela-
tive
decreases
in
myo-inositol
and
choline
in
the
cerebellum
(Singh
et
al.,
2011).
Given
that
a
functional
relationship
exists
between
PFC
and
cerebellum,
the
cerebellar
abnor-
malities
in
unaffected
relatives
at
risk
of
BD
have
been
proposed
to
be
resulted
from
a
compensatory
regulation
in
the
frontocerebellar
circuit
(Singh
et
al.,
2011).
Our
study
provided
more
direct
evidence
for
this
and
such
compen-
satory
up-regulation
of
GU
in
cerebellum
seemed
to
be
activated
by
insulin
as
a
result
of
the
decreased
GU
in
the
PFC.
Insulin
resistance
(i.e.,
reduced
insulin
sensitiv-
ity)
mostly
occurs
due
to
unhealthy
lifestyle
(Phielix
et
al.,
2011),
which
is
often
observed
in
patients
with
mood
dis-
orders
and
could
lead
to
fat
accumulation
in
the
peripheral
such
as
muscles,
which
in
turn
impedes
insulin
signaling,
affects
oxidative
capacity
of
muscular
mitochondria,
and
results
in
hyperglycemia.
The
observed
hyperglycemia
in
our
BD
pairs
may
be
also
resulted
from
an
increased
need
of
glucose
supply
in
the
‘brains
with
mitochondrial
dysfunc-
tion’,
in
addition
to
the
impaired
‘muscular’
mitochondrial
oxidative
capacity.
The
notion
of
an
increased
risk
of
insulin
resistance
in
those
with
mitochondrial
dysfunction
was
cor-
roborated
by
our
follow-up
results,
which
revealed
that
both
of
the
presence
of
risk
allele
(i.e.,
mtDNA-10398A)
and
higher
BMI
predicted
higher
HOMA-IR
six
months
later.
The
prevalence
of
diabetes
mellitus
was
reported
to
be
higher
in
BD
patients
than
in
the
general
population
(Cassidy
et
al.,
1999).
The
interpretation
of
our
findings
needs
to
be
tempered
by
some
limitations
of
this
study.
First,
the
potential
effects
from
medication
cannot
be
totally
removed.
However,
we
had
tried
to
avoid
the
confounding
from
medications
by
recruiting
un-medicated
patients
or
those
free
from
med-
ications
for
a
period
of
at
least
five
days.
Furthermore,
it
is
almost
impossible
to
recruit
euthymic
BD
patients
who
are
never
treated
by
medications
in
clinical
situations.
Sec-
ond,
the
link
between
mtDNA-10398A
polymorphism
and
BD
was
not
statistically
strong,
despite
its
significantly
higher
prevalence
among
the
BD
pairs.
However,
this
study
was
not
a
genetic
study
and
our
results
replicated
the
findings
that
mtDNA-10398A
may
represent
a
BD
risk
polymorphism
(Kazuno
et
al.,
2006,
2008).
Third,
MMSE
and
MoCA
were
not
frequently
used
in
non-dementia
populations.
Although
we
found
MoCA
sensitive
for
the
detection
of
mild
cognitive
impairment
in
the
BD
patients,
further
studies
are
needed
to
confirm
its
application
in
BD.
Finally,
although
the
difference
of
the
PFC
GU
between
the
single-stranded
mtDNA
subgroups
(i.e.,
10398A
vs.
10398G)
seemed
to
be
greatest
in
the
BDsib
(Fig.
2),
this
may
have
occurred
due
to
unbalanced
sample
sizes.
Unbalanced
sample
sizes
in
subgroups
defined
by
genetic
polymorphisms
are
almost
inevitable
in
genetic
imaging
studies
and
could
influence
imaging
statistics.
However,
to
what
degree
the
extent
of
the
between-genotype
difference
did
not
change
our
main
results,
since
all
groups
consistently
demonstrated
that
subjects
carrying
mtDNA-10398A
had
lower
GU
in
the
mPFC
and
dPFC
(Fig.
2),
which
are
energy-consuming
brain
regions
responsible
for
attention,
working
memory,
plan-
ning,
decision-making,
and
moderating
social
behaviors.
5.
Conclusion
This
first
study
provided
direct
evidence
to
support
that
mtDNA-10398A
had
functional
effects
on
brain
GU.
An
up-
regulation
of
glucose
utilizations
was
found
in
both
the
BD
and
BDs,
probably
compensating
for
mtDNA-10398A-related
energy
loss.
Role
of
the
funding
sources
The
study
was
sponsored
by
grants
from
the
Taipei
Veterans
General
Hospital
(V100E6-002,
V101E6-002,
and
V102E3-
006)
and
the
Ministry
of
Science
and
Technology
(MOST
102-2314-B-075-001
and
103-2314-B-075-072-MY3).
None
of
the
aforementioned
funding
organizations
had
any
role
in
the
study
design,
data
collection,
analysis,
interpretation
of
the
results,
writing
of
the
report,
or
the
ultimate
decision
to
submit
the
paper
for
publication.
Conflict
of
interest
statement
None
declared.
Acknowledgments
We
are
grateful
to
all
the
patients
who
participated
in
this
study
and
thank
Keresa
and
Aristine
Cheng
for
editing
English.
We
express
our
sincere
appreciation
to
the
indi-
viduals
in
the
Department
of
Nuclear
Medicine
of
Taipei
Veterans
General
Hospital
and
the
Brain
Research
Center
of
National
Yang-Ming
University
for
their
assistance
in
this
study.
Appendix
A.
Supplementary
data
Supplementary
data
associated
with
this
article
can
be
found,
in
the
online
version,
at
http://dx.doi.org/10.1016/
j.psyneuen.2015.02.003.
References
Arts,
B.,
Jabben,
N.,
Krabbendam,
L.,
van
Os,
J.,
2008.
Meta-
analyses
of
cognitive
functioning
in
euthymic
bipolar
patients
and
their
first-degree
relatives.
Psychol.
Med.
38,
771—785.
Bai,
Y.M.,
Chang,
C.J.,
Tsai,
S.Y.,
Chen,
Y.C.,
Hsiao,
M.C.,
Li,
C.T.,
Tu ,
P. ,
Chang,
S.W.,
Shen,
W.W. ,
Su,
T. P. ,
2013.
Taiwan
consensus
of
pharmacological
treatment
for
bipolar
disorder.
J.
Chin.
Med.
Assoc.
76,
547—556.
Balanza-Martinez,
V. ,
Rubio,
C.,
Selva-Vera,
G.,
Martinez-Aran,
A.,
Sanchez-Moreno,
J.,
Salazar-Fraile,
J.,
Vieta,
E.,
Tabares-
Seisdedos,
R.,
2008.
Neurocognitive
endophenotypes
(endophe-
nocognitypes)
from
studies
of
relatives
of
bipolar
disorder
subjects:
a
systematic
review.
Neurosci.
Biobehav.
Rev.
32,
1426—1438.

80
C.-T.
Li
et
al.
Baxter
Jr.,
L.R.,
Schwartz,
J.M.,
Phelps,
M.E.,
Mazziotta,
J.C.,
Guze,
B.H.,
Selin,
C.E.,
Gerner,
R.H.,
Sumida,
R.M.,
1989.
Reduction
of
prefrontal
cortex
glucose
metabolism
common
to
three
types
of
depression.
Arch.
Gen.
Psychiatry
46,
243—250.
Bingham,
E.M.,
Hopkins,
D.,
Smith,
D.,
Pernet,
A.,
Hallett,
W.,
Reed,
L.,
Marsden,
P.K.,
Amiel,
S.A.,
2002.
The
role
of
insulin
in
human
brain
glucose
metabolism:
an
18fluoro-deoxyglucose
positron
emission
tomography
study.
Diabetes
51,
3384—3390.
Bora,
E.,
Yucel,
M.,
Pantelis,
C.,
2009.
Cognitive
endophenotypes
of
bipolar
disorder:
a
meta-analysis
of
neuropsychological
deficits
in
euthymic
patients
and
their
first-degree
relatives.
J.
Affect.
Disord.
113,
1—20.
Buchsbaum,
M.S.,
Wu,
J.,
DeLisi,
L.E.,
Holcomb,
H.,
Kessler,
R.,
Johnson,
J.,
King,
A.C.,
Hazlett,
E.,
Langston,
K.,
Post,
R.M.,
1986.
Frontal
cortex
and
basal
ganglia
metabolic
rates
assessed
by
positron
emission
tomography
with
[18F]2-deoxyglucose
in
affective
illness.
J.
Affect.
Disord.
10,
137—152.
Cassidy,
F. ,
Ahearn,
E.,
Carroll,
B.J.,
1999.
Elevated
frequency
of
diabetes
mellitus
in
hospitalized
manic-depressive
patients.
Am.
J.
Psychiatry
156,
1417—1420.
Cumming,
T.B.,
Churilov,
L.,
Linden,
T. ,
Bernhardt,
J.,
2013.
Mon-
treal
Cognitive
Assessment
and
Mini-Mental
State
Examination
are
both
valid
cognitive
tools
in
stroke.
Acta
Neurol.
Scand.
128,
122—129.
de
Sousa,
R.T.,
Machado-Vieira,
R.,
Zarate
Jr.,
C.A.,
Manji,
H.K.,
2014.
Targeting
mitochondrially
mediated
plasticity
to
develop
improved
therapeutics
for
bipolar
disorder.
Expert
Opin.
Ther.
Targets
18,
1131—1147.
Folstein,
M.F.,
Folstein,
S.E.,
McHugh,
P.R.,
1975.
‘‘Mini-mental
state’’.
A
practical
method
for
grading
the
cognitive
state
of
patients
for
the
clinician.
J.
Psychiatr.
Res.
12,
189—198.
Gershon,
E.S.,
Hamovit,
J.,
Guroff,
J.J.,
Dibble,
E.,
Leckman,
J.F.,
Sceery,
W.,
Targum,
S.D.,
Nurnberger
Jr.,
J.I.,
Goldin,
L.R.,
Bun-
ney
Jr.,
W.E.,
1982.
A
family
study
of
schizoaffective,
bipolar
I,
bipolar
II,
unipolar,
and
normal
control
probands.
Arch.
Gen.
Psychiatry
39,
1157—1167.
Gispert,
J.D.,
Pascau,
J.,
Reig,
S.,
Martinez-Lazaro,
R.,
Molina,
V. ,
Garcia-Barreno,
P. ,
Desco,
M.,
2003.
Influence
of
the
normaliza-
tion
template
on
the
outcome
of
statistical
parametric
mapping
of
PET
scans.
Neuroimage
19,
601—612.
Heaton,
R.K.,
Chelune,
G.J.,
Talley,
J.L.,
Kay,
G.G.,
Gurtiss,
G.,
1993.
Wisconsin
Card
Sorting
Test
Manual.
Psychological
Assess-
ment
Resources.
Hirvonen,
J.,
Virtanen,
K.A.,
Nummenmaa,
L.,
Hannukainen,
J.C.,
Honka,
M.J.,
Bucci,
M.,
Nesterov,
S.V.,
Parkkola,
R.,
Rinne,
J.,
Iozzo,
P. ,
et
al.,
2011.
Effects
of
insulin
on
brain
glu-
cose
metabolism
in
impaired
glucose
tolerance.
Diabetes
60,
443—447.
Kato,
T. ,
Kato,
N.,
2000.
Mitochondrial
dysfunction
in
bipolar
disor-
der.
Bipolar
Disord.
2,
180—190.
Kato,
T. ,
Ishiwata,
M.,
Mori,
K.,
Washizuka,
S.,
Tajima,
O.,
Akiyama,
T. ,
Kato,
N.,
2003.
Mechanisms
of
altered
Ca2+
signalling
in
transformed
lymphoblastoid
cells
from
patients
with
bipolar
dis-
order.
Int.
J.
Neuropsychopharmacol.
6,
379—389.
Kato,
T. ,
2006.
The
role
of
mitochondrial
dysfunction
in
bipolar
disorder.
Drug
News
Perspect.
19,
597—602.
Kazuno,
A.A.,
Munakata,
K.,
Nagai,
T. ,
Shimozono,
S.,
Tanaka,
M.,
Yoneda,
M.,
Kato,
N.,
Miyawaki,
A.,
Kato,
T. ,
2006.
Identification
of
mitochondrial
DNA
polymorphisms
that
alter
mitochondrial
matrix
pH
and
intracellular
calcium
dynamics.
PLoS
Genet.
2,
e128.
Kazuno,
A.A.,
Munakata,
K.,
Kato,
N.,
Kato,
T. ,
2008.
Mitochondrial
DNA-dependent
effects
of
valproate
on
mitochondrial
calcium
levels
in
transmitochondrial
cybrids.
Int.
J.
Neuropsychophar-
macol.
11,
71—78.
Kempton,
M.J.,
Haldane,
M.,
Jogia,
J.,
Grasby,
P.M.,
Collier,
D.,
Frangou,
S.,
2009.
Dissociable
brain
structural
changes
associ-
ated
with
predisposition,
resilience,
and
disease
expression
in
bipolar
disorder.
J.
Neurosci.
29,
10863—10868.
Lee,
D.S.,
Kang,
H.,
Kim,
H.,
Park,
H.,
Oh,
J.S.,
Lee,
J.S.,
Lee,
M.C.,
2008.
Metabolic
connectivity
by
interregional
correlation
analy-
sis
using
statistical
parametric
mapping
(SPM)
and
FDG
brain
PET;
methodological
development
and
patterns
of
metabolic
connec-
tivity
in
adults.
Eur.
J.
Nucl.
Med.
Mol.
Imaging
35,
1681—1691.
Li,
C.T.,
Hsieh,
J.C.,
Wang,
S.J.,
Yang,
B.H.,
Bai,
Y.M.,
Lin,
W.C.,
Lan,
C.C.,
Su,
T. P. ,
2012.
Differential
relations
between
fronto-
limbic
metabolism
and
executive
function
in
patients
with
remitted
bipolar
I
and
bipolar
II
disorder.
Bipolar
Disord.
14,
831—842.
Matthews,
D.R.,
Hosker,
J.P.,
Rudenski,
A.S.,
Naylor,
B.A.,
Treacher,
D.F.,
Turner,
R.C.,
1985.
Homeostasis
model
assessment:
insulin
resistance
and
beta-cell
function
from
fasting
plasma
glucose
and
insulin
concentrations
in
man.
Diabetologia
28,
412—419.
McGuffin,
P. ,
Rijsdijk,
F. ,
Andrew,
M.,
Sham,
P. ,
Katz,
R.,
Cardno,
A.,
2003.
The
heritability
of
bipolar
affective
disorder
and
the
genetic
relationship
to
unipolar
depression.
Arch.
Gen.
Psychia-
try
60,
497—502.
Nasreddine,
Z.S.,
Phillips,
N.A.,
Bedirian,
V. ,
Charbonneau,
S.,
Whitehead,
V. ,
Collin,
I.,
Cummings,
J.L.,
Chertkow,
H.,
2005.
The
Montreal
Cognitive
Assessment,
MoCA:
a
brief
screening
tool
for
mild
cognitive
impairment.
J.
Am.
Geriatr.
Soc.
53,
695—699.
Phielix,
E.,
Szendroedi,
J.,
Roden,
M.,
2011.
Mitochondrial
function
and
insulin
resistance
during
aging:
a
mini-review.
Gerontology
57,
387—396.
Signorini,
M.,
Paulesu,
E.,
Friston,
K.,
Perani,
D.,
Colleluori,
A.,
Lucignani,
G.,
Grassi,
F. ,
Bettinardi,
V. ,
Frackowiak,
R.S.,
Fazio,
F. ,
1999.
Rapid
assessment
of
regional
cerebral
metabolic
abnor-
malities
in
single
subjects
with
quantitative
and
nonquantitative
[18F]FDG
PET:
a
clinical
validation
of
statistical
parametric
map-
ping.
Neuroimage
9,
63—80.
Singh,
M.K.,
Spielman,
D.,
Libby,
A.,
Adams,
E.,
Acquaye,
T. ,
Howe,
M.,
Kelley,
R.,
Reiss,
A.,
Chang,
K.D.,
2011.
Neurochemical
deficits
in
the
cerebellar
vermis
in
child
offspring
of
parents
with
bipolar
disorder.
Bipolar
Disord.
13,
189—197.
Stork,
C.,
Renshaw,
P. F. ,
2005.
Mitochondrial
dysfunction
in
bipo-
lar
disorder:
evidence
from
magnetic
resonance
spectroscopy
research.
Mol.
Psychiatry
10,
900—919.
Washizuka,
S.,
Ikeda,
A.,
Kato,
N.,
Kato,
T. ,
2003.
Possible
rela-
tionship
between
mitochondrial
DNA
polymorphisms
and
lithium
response
in
bipolar
disorder.
Int.
J.
Neuropsychopharmacol.
6,
421—424.
Willis,
M.W.,
Ketter,
T.A.,
Kimbrell,
T.A.,
George,
M.S.,
Herscov-
itch,
P. ,
Danielson,
A.L.,
Benson,
B.E.,
Post,
R.M.,
2002.
Age,
sex
and
laterality
effects
on
cerebral
glucose
metabolism
in
healthy
adults.
Psychiatry
Res.
114,
23—37.
Womer,
F. Y. ,
Wang,
F. ,
Chepenik,
L.G.,
Kalmar,
J.H.,
Spencer,
L.,
Edmiston,
E.,
Pittman,
B.P.,
Constable,
R.T.,
Papademetris,
X.,
Blumberg,
H.P.,
2009.
Sexually
dimorphic
features
of
vermis
morphology
in
bipolar
disorder.
Bipolar
Disord.
11,
753—758.
Young,
S.E.,
Mainous
3rd.,
A.G.,
Carnemolla,
M.,
2006.
Hyperinsu-
linemia
and
cognitive
decline
in
a
middle-aged
cohort.
Diabetes
Care
29,
2688—2693.